Anexo 2. Manuales de Perfil por Competencias
2. Jefe Administrativo Financiero
A number of factors should be taken into account in considering the findings of this study. First, much of the data are collected by the service providers. The outcome measures available from CMS, shown in Tables 3 and 10 above, relate to resident and patient characteristics. For the most part, these measures require subjective determinations of resident and patient attributes by nursing home and home health care staff, determinations such as whether residents’ “need for help with daily activities has increased,” or whether a patient has “gotten better at getting in and out of bed.” These determinations are subject to variation in observational skills, biases, and understanding of terminology used in the reporting process.
The CMS data set only includes data on nursing homes and home health agencies that are Medicare/Medicaid certified. Other providers are not included in the data or data analysis. Differences that might have become apparent if all service providers were compared, regardless of Medicare/Medicaid certification, could not be determined.
Information about the accuracy of the data in the nursing home data set is included in a report by the Department of Health and Human Services Office of the Inspector General.
We found that Nursing Home Compare contains nearly all Medicare- and Medicaid-certified nursing homes. However, one or more surveys were missing from 19 percent of nursing homes, leaving consumers with incomplete information about those homes’ survey and complaint histories. Inspection results on Nursing Home Compare are largely accurate, but one or more deficiencies were missing from 11 percent of nursing homes’ inspection results, and Nursing Home Compare presents deficiencies not found in State survey documentation for 15 percent of nursing homes. Inaccuracies may be due to late data entry by State survey agencies, no tracking of inaccuracy by CMS, and failure of State survey agencies to transmit data on amended deficiencies.20
The nursing home and home health agency data sets did not include complete data for a significant number of service providers. The information below details the percentages of missing data for resident and patient characteristics.
20 Inspection Results on Nursing Home Compare: Completeness and Accuracy, Department of Health and Human
Services Office of the Inspector General, June, 2004, at http://www.healthlawyers.org/docs/ask2004/ OEI_01_03_00130.pdf.
The Roundtable on Religion and Social Welfare Policy 31 Nursing Home Data
The percentages of nursing homes for which data were included in the CMS data set for the fourteen resident characteristics are listed below. Percentages were calculated based on data for each of the characteristics. The CMS data set includes two reasons that data were missing: “The number is too small to report” and “The data for this outcome is missing.”
Nursing Home Type
Percentage of All Data Elements with Valid Data
All Nursing Homes 72.89%
Church-Related Nursing Homes 68.89%
All Non-Religious Nursing Homes 73.12%
Other Non-Profit Nursing Homes 64.78%
Home Health Agency Data
The percentages of home health agencies for which data were included in the CMS data set for the eleven patient characteristics are listed below. Percentages were calculated based on data for each of the characteristics. When data on patient characteristics were missing, the data set indicates that “This agency currently does not have data for this measure or this agency has less than 6 months of data.”
Home Health Agency Type
Percentage of All Data Elements with Valid Data
All Home Health Agencies 76.69%
Religiously-Affiliated Agencies 91.41%
All Non-Religious Agencies 75.71%
Other Non-Profit Agencies 87.27%
Another issue was reliability of the identification of faith-affiliated service providers. We found that there was significant variation between the identification of type of ownership in the CMS nursing home data and identification by field researchers, though this did not greatly affect the findings. A related issue involved terminology. A number of undefined terms are used to identify service providers with a connection to a religious organization. For example, CMS uses at least two different terms – church-related and ownership with a religious affiliation – though it does not appear that the different terms are intended to have different meanings. No definition of these terms could be found
The Roundtable on Religion and Social Welfare Policy 32
after numerous contacts with CMS staff and an examination of the forms from which data regarding the type of provider are drawn.
Only portions of the performance data collected for nursing homes and home health agencies are available on the CMS website and in the data sets that CMS makes available for research purposes. For example, CMS includes only 11 of the 41 data elements related to home health agency performance in the Home Health Compare utility. It is possible that other data might have affected the findings.
Another concern related to the influence that the conditions of the residents and patients could have on outcome measures. For example, a home health care agency that accepts unusually difficult patients might have poorer outcomes than other agencies. However, this should not be a significant factor influencing the findings because CMS adjusts for patient risk for home health care patients through a statistical technique that “accounts for differences in the agency's patients versus the reference sample, and minimizes the possibility that the differences are due to factors other than the care provided by the agency.”21 A number of the nursing home resident characteristics are also risk-adjusted: “To reduce the chance that a nursing home that serves more frail residents appears worse due to its resident population, certain residents are not included in the calculation of a quality measure. This makes the resident population used to calculate the quality measures more similar, therefore allowing comparison between nursing homes on these measures.”22
Two additional cautions are also appropriate. Because the data do not include any information about the religious character of services or service providers, any conclusions regarding comparative performance cannot be directly attributed to such factors. In the findings above, we only report what the data indicated regarding the various performance measures included in the data sets. We do not attribute the differences to particular characteristics of types of service providers. The second caution is that the findings relate specifically to the types of service providers included in the data sets – nursing homes and home health agencies – and should not be construed as being applicable to all service providers.
Notwithstanding these limitations, the data sets proved useful for the purposes of this study – to determine whether available administrative data could shed light on the question of the relative effectiveness of services provided by faith-affiliated and secular service providers, and if so, what those data might show about comparative performance.
21 From the Medicare Quality Improvement Community website at http://www.medqic.org/content/nationalpriorities
/topics/projectdes.jsp?topicID=417&pageID=3#measures.
The Roundtable on Religion and Social Welfare Policy
www.ReligionandSocialPolicy.org(518) 443-5014